黏性边界层网格自动生成

AUTOMATIC VISCOUS BOUNDARY LAYER MESH GENERATION

高雷诺数黏性流动在壁面附近存在边界层,在计算模拟中自动生成可靠且有效的计算网格仍然是计算流体力学存在的瓶颈.三棱柱/四面体混合网格技术在一定程度上缓解了这个困难.然而,对于复杂外形的情况,在边界层内自动高效生成高质量的三棱柱单元仍然十分困难.常用的层推进法在凹凸区域及角点处生成的边界层网格单元质量较差,边界层网格最外层尺寸不均匀.针对这些问题,发展了一种黏性边界层网格自动生成方法,通过顶点周围边的二面角识别物面网格特征确定多生长方向,预估并调整生长高度处理相交情况.同时提出一种三维前沿尺寸调节方式,提高了边界层网格单元的正交性,保证了边界层网格与远场网格尺寸的光滑过渡.通过ONERA M6翼型以及带发动机短舱的DLR-F6翼身组合体等外形的网格生成实例及绕流数值模拟,将计算值与标准实验值进行对比,结果表明:该方法能够自动高效地生成满足数值计算需求的混合网格.

High Reynolds number fluid flows have boundary layers at the wall.To automatically generate robustand valid boundary layer mesh for the simulations is still the bottleneck problem of computational fluid dynamics.Prisms/Tetrahedra hybrid mesh leads to significant savings in mesh size and solution costs.However,it’s still diffi-cult to generate prismatic elements of high quality within boundary layers of complex models.Previous advancing layerstechniques sometimes lead to invalid meshes and poor quality elements at concave/convex ridges and sharp corners.Toimprove these situations,we present a strategy for automatically generating viscous boundary layer mesh.In this method,multiple growth directions at ridges and corners are well defined by the dihedral angles around the vertices and the growthheights are adjusted appropriately.Therefore,boundary layer mesh grows well at sharp corners,convex and concaveridges of the domain.We also decrease the number of global intersection checks by predefining the total growth heightsbefore generating elements through one global check,which improves the efficiency of mesh generation.At the sametime,we develop a3D strategy of mesh size control to get a size uniform triangular mesh of the outer boundary of theboundary layer mesh,which is beneficial to generate far-field isotropic mesh of high quality.Finally,mesh examples andthe viscous flow simulations including2D and3D are presented.In3D,the hybrid mesh over the standard ONERA M6and DLR-F6configurations are generated with the present method.The numerical results agree very well with experiment data which indicates that the hybrid viscous meshes generated by the proposed method are effective and efficient